Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: THEend8_
COMP90086 Computer Vision
Submission: Source code and written report (as .pdf)
Marks: The assignment will be marked out of 30 points, and will contribute 30% of your
total mark.
Modern computer vision algorithms frequently meet or exceed human performance on constrained
supervised tasks like object classification or face recognition. However, there are still many gaps
between human and AI performance. In particular, humans are better at tasks that require flexible and
abstract reasoning about images. One task that has been proposed to evaluate human-like perception
of images is the Totally-Looks-Like challenge [1]. This task is based on a popular entertainment
website where users share pairs of
images of things that they think look similar, such as the example show in Figure 1.
In this project, you will develop an algorithm to solve the Totally-Looks-Like challenge. Your algo-
rithm will take one image from a Totally-Looks-Like pair as input and attempt to find its match from
a list of possible candidates. This task is challenging because this dataset reflects many different types
of image similarity – two images may be paired because they contain similar colours, shapes, textures,
poses, or facial expressions. Sometimes only part of the image is relevant to the comparison. You may
need to consider a variety of different features of each image to find the best match.
Whatever methods you choose, you are expected to evaluate these methods using the provided data,
to critically analyse the results, and to justify your design choices in your final report. Your evaluation
should include error analysis, where you attempt to understand where your method works well and
where it fails.
You are encouraged to use existing computer vision libraries your implementation. You may also use
existing models or pretrained features as part of your implementation. However, your method should
be your own; you may not simply submit an existing model for this problem.